In February 1998, Jesus edwardsii larval catches off the east coast of New
Zealand were binomial in size and stage of development. Mid-stage larvae (p
resumed to be from the 1997 cohort) appeared to be contained within a large
permanent eddy found offshore, known as the Wairarapa Eddy. Late-stage lar
vae (1996 cohort) have a different distribution from the mid-stage larvae,
being found closer inshore of the eddy. Pueruli (also 1996 cohort) were fou
nd well inshore of the eddy. Surface currents derived from TOPEX/Poseidon a
ltimeter measurements of sea level are used to simulate larval concentratio
ns. The numerical simulations show that the different distributions of the
3 groups cannot be explained on the basis of advection alone. Mid-stage con
centrations appear to be well described by passive drift alone, but the sim
ulations suggest some mechanism inducing shoreward transport is needed for
both pueruli and late-stage phyllosomas. Observed puerulus concentrations a
re best matched by adding a shorewards-directed speed of 8 to 10 cm s(-1) f
or between 1 and 3 wk before the date of the cruise. Late-stage phyllosoma
concentrations are best matched by adding speeds of 4 to 6 cm s(-1). Sustai
ned shorewards swimming speeds of 8 to 10 cm s(-1) are well within the esti
mates of swimming speed that have been reported elsewhere for pueruli. To o
ur knowledge, extended directed horizontal swimming has not been reported f
or phyllosomas, yet in the final stage, the pleopods are almost as well dev
eloped as they are in the pueruli. These final stage larvae could well use
forward pleopod propulsion, in conjunction with exopodal setai beating, to
allow them to swim.